The statements in this section merely provide background information related to the present disclosure and may not constitute prior art.
Many technologies have been applied to networks of Ethernet technology, e.g., Provider Backbone Transport (PBT) is deployed based on Provider Backbone Bridges Network which is defined by IEEE 802.1ah (i.e., MACinMAC).
IEEE 802.1ah defines a new architecture and bridge protocol which are compatible and interoperable with IEEE 802.1ad Provider Bridge (PB). Therefore multiple PB networks can be connected and form at least 224 Virtual Local Area Networks (VLAN).
The structure of the Provider Backbone Bridges Network defined by IEEE 802.1ah is shown in FIG. 1. The Provider Backbone Bridges network is built on the basis of IEEE 802.1ad (also called QinQ) and is able to encapsulate a complete PB packet into a Provider Backbone Bridge (PBB) packet. Therefore, the Provider Backbone Bridges network provides a hierarchy network, and serves as the basis of Ethernet connection or tunnel technology.
A message encapsulated as defined in IEEE 802.1ah is shown in table 1:
TABLE 1B-DAB-SAB-TagI-TagS-TagC-DAC-SAC-DATA
The B-DA is a Provider Backbone Bridge Destination MAC Address, the B-SA is a Provider Backbone Bridge Source MAC Address, the B-Tag is a Provider Backbone Bridge Tag, the I-Tag is a Service Instance Tag, the S-Tag is a Service Tag, the C-DA is a Customer Destination MAC Address, the C-SA is a Customer Source MAC Address, and the C-Data is Customer Data.
As can be seen from table 1, a customer packet, transported from the PB to the PBB and including the C-DA, C-SA and C-Data, is completely encapsulated in the PBB packet. The Provider Backbone Bridge MAC Address (B-MAC) (i.e., the B-DA and B-SA) is the MAC address of PBB device. The B-Tag is defined on the basis of IEEE 802.1Q. In a PBB network, a packet may only be forwarded in a standard Ethernet forwarding process on the basis of the B-MAC and the B-Tag according to IEEE 802.1Q.
It can also be seen from table 1 that in this hierarchy network, the MAC address of the customer is isolated from the MAC address of the provider network device. The TAG of the customer data is isolated from the TAG of the provider TAG.
The PBT technology on the basis of the IEEE 802.1ah provides features as follows. A PBT device needs to support Individual VLAN Learning (IVL). Because the PBT device supports the IVL, VLANs may be categorized into connection-oriented VLAN (i.e., PBT VLAN) and a connectionless VLAN (i.e., common VLAN). In other words, a part of the VLANs may be designated as the PBT VLAN on PBT related device and the part of VLANs can function independently without influencing common VLANs, MAC address learning and spanning tree protocol are disabled in the PBT VLAN, and broadcast function as well as multicast and broadcast of unknown packets in the PBT VLAN are disabled.
The PBT network is shown in FIG. 2. The provisioning and management system of the network connects with all Provider Edge Bridges (PEB, the PEs in the drawing) and Provider Bridge (PBs, the Ps in the drawing) in the network and is used for configuration and link maintenance, e.g., control and management functions such as status detection and path protection.
As shown in FIG. 2, the PBT network includes a number of PEs and Ps which are generally Ethernet switches supporting IVL. The MAC addresses of destination devices and the PBT Virtual LAN Identifiers (VID) are configured respectively in the Ethernet switches. The destination MAC address and the VID constitute a tag which is an identifier of a channel. An Ethernet Switched Path (ESP) is formed by transporting the identifier of the channel over the plurality of the Ethernet switches supporting the IVL. The channel may be a connection, i.e. Ethernet Virtual Connection (EVC) or tunnel. The MAC address, VID and forwarding mode used by the PBT technology are in compliance with Standard IEEE 802.1q.
The method for forwarding a data packet is as follows.
The provisioning and management system configures links, for example, configures the links of the PBT dynamically or statically via the control plane of General Multi-Protocol Label Switching (GMPLS), and maintains the state of the links.
Subsequently, the provisioning and management system learns the MAC address by using the IVL on the network nodes, such as the PE and P in FIG. 2, on the transport path ESP of the PBT, and configures the forwarding table of MAC addresses of the PBT.
A PE forwards customer traffic to a next-hop node according to the forwarding table; the next-hop node forwards the customer traffic to another next-hop node according to the forwarding table. At last, the customer traffic is forwarded to another PE connected with a destination customer network. The intermediate network nodes, i.e., the Ps, forward packets without any other processing.
For example, in FIG. 2, the provisioning and management system configures an ESP from PE1 to PE3 statically. A forwarding table is configured on PE1, intermediate Ps and PE3. The MAC address of PE3 and VLAN 44 are taken as the tag of the ESP. The MAC address of PE1 contained in packets would indicate the source device of the ESP. The ESP is a unidirectional path.
FIG. 3 is a schematic diagram illustrating the structure of a PBT network node device in the prior art. The device includes a first module, a second module and a third module. The third module is optional. The first module may be a physical layer processing module mainly used for providing physical layer functions. The second module may include a MAC layer processing unit and a forwarding unit. The MAC layer processing unit is used for providing MAC layer processing functions including IVL, MAC address configuration and MAC address forwarding table maintenance. The forwarding unit is used for forwarding, via the first module, the data packet processed on the MAC layer. The third module may include a data processing module of layers above layer-2. The MAC layer processing unit of the second module sends, according to the information in data packets such as the destination address, processed data of data packets to the third module for data processing on layers above layer-2. The data processed by the third module is sent to the first module through the forwarding unit in the second module and is further forwarded to other network nodes. The MAC addresses in the prior art are globally unique MAC addresses of devices. Therefore the second module has to perform IVL to learn the MAC address of other PBT network node device to configure and maintain the MAC address forwarding table.
It can be seen that the PBT technology in the prior art has the following disadvantages.
The MAC addresses in the prior art are globally unique MAC addresses of devices which can not be planned by a network administrator in a unified manner. Therefore, MAC address convergence can not be implemented, and location of the connection ends can not be learned through MAC addresses.
Because the length of a MAC address plus VID is 60 bits, theoretically a label containing a combination of a MAC address and a VID may have 260 possibilities of values. However, the MAC addresses actually in use are MAC addresses of devices, and the number of ESPs that reach a device is determined according to the allocated PBT VID which is only 12 bits long and provides only 4096 possibilities, so the tag is subject to narrow choices.
The ESP is identified with the MAC address of the destination device and the VID, the destination MAC address is the global MAC address of the device and a packet is forwarded by intermediate nodes to one single destination node instead of multiple destination nodes.